Guide to ethical research and design practice when working with vulnerable participants and on sensitive topics

Abstract

The purpose of this paper is to establish a set of guidelines for student designers and design researchers when working with vulnerable participants or/and on sensitive topics. This paper explores several case studies of design research projects where vulnerable participants have been involved at various stages and to varying degrees. The case study analyses key questions such as when is it/ not appropriate to involve participants? What are the most useful methods to work with participants? When are alternative methods of research and testing sufficient? How can expectations be managed? And what is the payback for people to participate? The paper concludes by collating the key lessons into a ‘living’ guide to which design practitioners, design educators and researchers can refer.

Keywords:

• design research
• ethics in design
• vulnerable participants in design
• sensitive topics in design
• design education

Highlights

• Ethical practice in design research is complex due to the evolving and iterative nature of the requirements of participants.

• A living guide is presented offering direction on ensuring ethical participant involvement in design research.

• Where it is unethical to involve a vulnerable participant, alternative methods and workarounds are proposed.

Introduction

Current research would advocate that design research should be conducted with end-user groups and key stakeholders to ensure that solutions developed meet the needs and expectations of those most impacted (Merter and Hasırcı 2018; Sanders and Stappers 2008). The benefits of co-designing with vulnerable users and stakeholders are that they can become empowered participants who can influence the path to the creation of solutions that address their needs. As designing with vulnerable users and on sensitive topics becomes more prevalent, there is a clear need to establish societal and policy guidelines to ensure ethical practices protect both participants and designers (Mulvale, Moll, Miatello, Robert, et al. 2019). Along with adherence to standard ethical requirements when conducting research with vulnerable groups and dealing with sensitive topics, there are often additional considerations when developing design solutions. Design research and practice projects present different ethical challenges to other disciplines including the extent of and expectations from participant involvement; dynamic participant roles (e.g. from users to co-designers); changing engagement methods; expected versus actual project outputs and project durations. Involving people may not always be ethical or appropriate, for example, in student projects or in projects where implementation is unlikely.

The impetus for these Research Notes stems from the researchers’ experiences working with design students at the University of Limerick and observing the disconnect between this and the emergent need for ethical design practices. While Sanders and Stappers (2008) and Shore et al. (2018) would advocate for co-design and participatory design research, Mulvale, Moll, Miatello, Robert, et al. (2019) state that a better understanding is needed of how to facilitate the involvement of vulnerable populations in participatory design research to avoid reinforcing existing inequities. They highlight several challenges to the participation of vulnerable participants: engagement; power differentials; health concerns; funding; and other economic/social circumstances.

While there is an acknowledgement of the value of design in these settings, concerns are also being expressed as to whether design researchers and students are adequately equipped to guide complex interactions with vulnerable participants (Craig et al. 2021). When participants are viewed as vulnerable, there are concerns that researchers may not have the ability to act appropriately due to power differentials between the researcher and participant and may themselves be vulnerable when exposed to emotional encounters that are upsetting and difficult to deal with (von Benzon and van Blerk 2017). There is a potential risk of harm to both parties if the process is not managed carefully.

Many vulnerable groups may initially be very excited at the prospect of designing solutions that can improve their quality of life. They may be willing to engage in projects as research participants or as co-designers across the various stages. However, there is a risk that these participants may invest their time, knowledge, and expertise with great expectations but end up with little in return (Suri 2015). A major ethical concern is that they may feel used, exploited, and let down as projects do not reach full completion. Including vulnerable participants in design research, projects must be beneficial to participants beyond the goals of the project; otherwise, alternative methods should be employed. There is the need to recognise situations where participant involvement may not be necessary, and if this is the case, continuously ensuring rigour and validity of the design process is paramount. Therefore, the aim of these notes is to explore how a ‘living’ guide can be developed on the appropriate involvement of vulnerable participants possibly on sensitive topics across design research projects and methods that can be employed as alternatives to involving the targeted user group.

This paper explores several case studies in academia from undergraduate to postdoctoral levels where designers have worked with vulnerable participants and on sensitive topics across different stages of the design process. The case studies are analysed to uncover the successful ways in which the designers engaged with and involved participants in their design process. Also explored are the ways in which this participation could have been improved or even removed if the burden of participation became overly complex or time-consuming with little reward for the people involved.

Further objectives of the notes are to

1. Create a set of protocols to determine if and at what stages of projects vulnerable participants should be involved.

2. Propose an ethical living guide for involving participants in design research projects, where and when appropriate.

3. Propose alternative means or guides to working with vulnerable end-user groups in design research projects.

1. Background

1.1 Need for ethics in design

The consideration of ethics in design is relatively recent and has come about due to the broadening use of increasingly pervasive digital technologies (Franssen, Lokhorst, and Van de Poel 2009; Gabel, Schobert, and Krüger 2024; Steen 2015). Until recently, it was considered that new technologies created new capabilities for people and positive experiences, therefore not requiring ethical consideration. New design technologies were held by a neutral view, which posits that technology is a neutral instrument that can only be put to good or bad use by its users. This position has been critiqued robustly with the argument that through the device or solution it creates, the design will guide and steer behaviour, showing that it cannot be value-neutral (Van de Poel and Kroes 2013). Therefore, designer researchers need to question how responsible their practices are and how risks and potential unintended consequences of design technologies can be avoided (Franssen, Lokhorst, and Van de Poel 2009).

Involving users and experts as participants in design research is seen as an essential part of the design process to create relevant and necessary solutions (Sanders and Stappers 2008). Every design research project must be built on a robust ethical foundation and conducted with integrity and rigour (Cash et al. 2022; Craig et al. 2021). However, when people are involved in a participatory capacity in design research (as experts, users, co-designers and other key stakeholders), the parameters of the project are often dynamic and changeable (Munteanu et al. 2015; van Oorschot et al. 2022). Both experts and users may be classified as vulnerable, and, on this premise, the standard ethics procedure within academic institutions might not be nuanced enough or arguably too ‘rigid’ to cover the diversity of activities or user involvements across the entire design process with the dynamic needs of the participants often overlooked (Dewing 2008; Munteanu et al. 2015). A participant can be involved in more than the initial primary research stage, having a repeated role in the subsequent stages of ideation, concept development and even implementation across a practice-based design and research project. Conversely, the ethical approval process might be protracted and overly complex for shorter UG and PG student projects (Godbold, Lees, and Reay 2019). As such, the role of the participant needs a clearer definition and factors impacting and affecting their involvement must be considered carefully at the outset of any design research project.

Design research engages with participants over different stages of the project process. Figure 1 visually outlines this process and identifies the points of engagement with research participants. While the process is presented linearly in Figure 1, we acknowledge that this process is iterative as designers may need to go back and forth between stages (Chiapello and Bousbaci 2022; Lawson and Dorst 2013). As well as responding to needs findings at the early research phase, participants may also be asked to actively test proposed treatments or remedies while using prototypes at various levels of fidelity repeatedly over time. Chiarini cited in Herwix et al. (2022) states that even good design can have adverse effects and that design researchers must consider the ethical consequences of their actions at all steps of the design process. In addition, the medical literature advises that exceptional care be taken when recruiting patients at a vulnerable time in their lives and often the use of proxies is preferable instead (Reid and Brief 2009).

Figure 1. Design process and points of engagement for research with participants.

1.2. Vulnerable participants

Working with users as participants introduces a complexity to design research and this is further amplified when the participants are classified as vulnerable. While there is no consistent definition of a vulnerable participant in a research or design context available across international policies (Bracken-Roche et al. 2017), vulnerable participants occupy a diverse and evolving categorisation and currently include racial and ethnic minorities, people with additional physical or cognitive needs, persons who are incarcerated or in unequal or dependent relationships, as well as elderly individuals, youths and children (Rios 2016). Vulnerability is individual and context-dependent, and everyone may be impacted by different and often multiple vulnerabilities (GOV.UK 2023). The NHS (2023) defines a vulnerable adult as ‘any adult (person over the age of 18) unable to take care of themselves or protect themselves from exploitation’ (ibid) whilst the HSE (2023) offers the definition of a person who ‘may have difficulty and need support in making decisions’. These vulnerabilities thus may affect prospective participants’ ability to understand, consent and participate in design research (ibid). As children and young people are potentially vulnerable participants, their rights as encapsulated in the UN Convention of the Rights of the Child must be protected, particularly with reference to Article 13 that addresses the right to seek, receive and impart information (Unicef 2007).

The inclusion of vulnerable participants and the exploration of sensitive topics can add perceived complexity to a project, and there is a growing recognition of the value of designers and design approaches in these contexts (Craig et al. 2021) From design-led projects on autism spectrum disorder (Merter and Hasırcı 2018), homelessness and drug use (Sellen et al. 2022) to menstrual health (Flemings et al. 2018; Fox 2018) and menopause (Tutia et al. 2019), we can see the emergence of design research as a tool for understanding complex situations and a process for generating solutions. However, concerns are raising amongst the design community that design researchers may not be adequately equipped to work in such sensitive ethically complex areas (Munteanu et al. 2015). They found that researchers were making decisions based on personal moral principles rather than on clear guidelines. They propose recommendations for a situational approach to ethical issues in the planning and execution stages of the research. Craig et al. (2021) supported this position and developed an ethical Roadmap to facilitate discussion and be a form of rehearsal for potentially sensitive ethical topics and dilemmas. The roadmap encourages a reflective and reflexive process to build self-awareness of how a researcher’s values are brought to bear during the research project.

There are also several concerns and challenges when vulnerable people participate in design research (Mulvale, Moll, Miatello, Robert, et al. 2019). Some of these may be around issues with the initial identification and recruitment of participants and the availability of these participants to engage repeatedly particularly due to health conditions or changing circumstances. Therefore, the inclusion criteria for a person to be able to participate in the research must be considered. Johansson (2019) argues that in planning the study, the researcher must consider how someone, for example, with an impairment, a mental health issue or someone who is homeless can take part in the research, and the researcher may be required to put in place accessible methods and toolkits to enable participation.

Power differentials during the participatory co-design phase are also a concern (Sendra 2022), and funding challenges may also arise to support the implementation of the projects. Power differentials may occur when there are significant differences in educational and cultural backgrounds, language barriers, and gender imbalances (Jiang et al. 2022). This can occur between the researcher and the participant or between different participants or stakeholders. Vulnerability, therefore, must be viewed beyond an individual’s ability to consent, but more as a relational factor stemming from power imbalances, potential for harm, cognitive ability, interpersonal relationships, cultural variances, and social imbalances (Godbold, Lees, and Reay 2019).

Exclusion from participation does not serve either the person or the design process as key insights, needs and perspectives may be missing, and the final outcomes not fit for purpose as a result (Kanstrup and Bertelsen 2016). The broadening of inclusion factors will ensure better and more responsible engagement with users and create the potential for an elevated level of trust with participants to continue to engage in design research and testing. Therefore, every effort should be made to include eventual users as participants in design research.

There are different levels also at which participants can engage. Hart (2013) devised ‘the Ladder of Youth Participation’, a modelling tool that describes nine possible levels of youth participation in the context of research projects. These nine levels range from non-participation or tokenistic involvement to participation in the form of consultation and informing the participant, to the top level, which sees the participants leading and sharing in the decision-making process. Therefore, consideration needs to be given to the appropriate level of engagement of all vulnerable participants.

1.3. How to do ethical design research and practice

Several ethical approaches have been developed within design, as outlined in Table 1. These range from considering moral values with the designed output and that it does not negatively impact on citizens to different levels of participation and ownership by stakeholders during the research and design phase. Value-sensitive design (VSD) is one approach that accounts for human values throughout the design process (Friedman, Hendry, and Borning 2017). The focus of value-sensitive design is to incorporate human values in the design process. Further approaches that address similar concerns have been developed (Huldtgren 2015; Snyder, Shilton, and Anderson 2016), including values in design (Nissenbaum 2001), values for design (Van den Hoven, Vermaas, and Van de Poel 2015) and worth centred design (Cockton et al. 2009). There has been criticism, however, that these approaches do not commit to a particular ethical theory and, therefore, lack the means for gathering moral values from stakeholders as opposed to preferences on the features of a designed solution, rendering it unprincipled (Jacobs and Huldtgren 2021).

Table 1. Research approaches in design.

Research approach in design	Description
Value Sensitive Design (VSD) (Friedman, Hendry, and Borning 2017). Values in Design (Nissenbaum 2001), Values for Design (Van den Hoven, Vermaas, and Van de Poel 2015), Worth Centred Design (Cockton et al. 2009)	Accounts for human values within the final solution and technology. Accounts for direct and indirect stakeholders. Participants are engaged as research subjects/participants throughout the design process
Participatory Design (Borning, Friedman, and Kahn 2004)	Engages users and stakeholders in both the research and design process and strives for equal partnership with equal power.
Co-Design (Sanders and Stappers 2008) Experience-based co-design (Mulvale, Moll, Miatello, Robert, et al. 2019)	Engages users and stakeholders in both the research and design process as co-designers and strives for equal partnership with equal power.
Situated Ethics (Ransom et al. 2023). In-Action ethics (Frauenberger, Rauhala, and Fitzpatrick 2017)	Inclusionary framework to ensure symmetrical relationships and equal power balances. Ensuring responsiveness as situation arise. Reflexive approach.

Other approaches have recognised the importance of including the needs of the people impacted by the design outcome and participatory design engages users in both the research and design process (Bjerknes and Bratteteig 1995). Participatory design recognises people’s right to participate in shaping the products and services that they will use (Donetto et al. 2015) with an emphasises on equal partnership with equal power between the designers and participants (Borning, Friedman, and Kahn 2004). Co-design goes further where stakeholders not trained in design work with designers to develop solutions. Co-design brings together designers, users, novices, experts, citizens, or customers together to design collaboratively and deems that everybody is an expert regarding their own experience (Cavignaux-Bros and Cristol 2020). Mulvale, Moll, Miatello, Robert, et al. (2019) applied an experience-based co-design (EBCD) methodology to work directly with the people affected by a situation (patient, user, etc.) in partnership with designers and other stakeholders to build understanding and then develop solutions to improve the situation in a healthcare setting. They recommend a suite of tools and techniques to involve vulnerable users fully in the process across all stages  – experience mapping, card sorting and cognitive walkthroughs, developing a common language to describe experiences and help with understanding (ibid).

Whilst ethics training for design researchers may be lacking, piecemeal and variable (Craig et al. 2021) much can be learned from other projects where vulnerable participants and/or sensitive topics are explored. Designers may need to go further to safeguard the inclusion of vulnerable participants and safeguard them in the research and design process. Development Ethics and situational ethics, for an example, posit an inclusionary framework to ensure symmetrical relationships and equal power balances that retain a systems perspective and acknowledge the interconnectedness of issues and user expectations (Ransom et al. 2023). Applying this approach allows designers to respond to situations, problems and complexities that may emerge during a project as roles shift and mechanisms of engagement evolve. Building on the theory of situational ethics, Frauenberger, Rauhala, and Fitzpatrick (2017) advocated for in-action ethics where reflection in action allows for a reflexive approach to a situation as the dynamics and situations change. They argue that current models of ethical approval are too static, anticipatory, and formalised (ibid).

Further work by Mulvale, Moll, Miatello, Murray-Leung, et al. (2019) defines a framework to address some of the challenges for co-designing health and social services with vulnerable populations. They propose that the recruitment of participants must be built on trust, requiring flexibility and responsiveness. The engagement of participants requires user-centeredness and reciprocity, where power differentials can be addressed by empowerment. Power sharing and health concerns can be approached by creating a safe process and a conducive environment and by recognising emotional toil. They state that it is important to respect cultural differences to recognise the person in context and finally that acquiring funding for these projects requires building credibility, demonstrating impact and being opportunistic (ibid).

Similarly, Kanstrup and Bertelsen (2016) recommend meeting the participants where they are at, in locations and environments where they are comfortable, and at the level at which they engage (e.g. wearing similar clothing, language used, and communication modes employed, etc.). Using straightforward language, legitimising the opinions of the participants by ensuring they are heard, and their opinions/perspectives/feedback considered as the project progresses, all positively contribute to ethical practices. On this premise, Larkin, Boden, and Newton (2015) caution that managing expectations from the outset can diffuse tensions, acknowledge the diversity of perspectives and ensure any changes do not negatively impact on the project. These serve to support symmetrical relationships and even power balances between researcher and participant (Munteanu et al. 2015).

A toolkit developed by Glipsen (2017) comprises exercises to help designers identify the values they can embed into their designed outputs (tangible products, digital products and services). The various activities across the toolkit explore how designed outputs can meet the often-conflicting concerns of the stakeholders, who in turn should be afforded agency over the outcomes. Lettis et al. (2020) place a focus on the personal values and ethics of designers themselves and how these influence designers developing practices and professional decisions.

Craig et al.’s (2021) Ethical Roadmap, through its seven components/ activities, moves the discussion forward for design-led research projects in responsive and practical ways. Activities range from aligning the values of the full project team, identifying roles, to drawing in experts and critical friends as a project evolves, as well as tools to support the framing of initial consent and the management of ongoing consent. Most interestingly are the provocations and the Warp & Weft activities. The former presents potential ethical issues for discussion, which will enable teams to foreground and identify problems in a planned project. The latter (Warp & Weft) activity invites all stakeholders to voice their wants and don’t want from the project thus surfacing the expectations of all project members, so processes and protocols can be negotiated and agreed upon. This flexible and easily understood roadmap helps in preparing and laying the groundwork for the implementation of ethical and productive design and Human–Computer Interaction (HCI) research projects.

Suri (2015) advises on how to conduct ethical research during the ‘dynamic and edgy’ design process. Starting with how to approach and build a relationship with people to safeguarding interactions over the full design process, the guide steps through each stage of the design process from planning to prototype testing and eventual data sharing/ use. Communication and transparency around research activities and managing expectations as well as sensitivity in dealing with everyone involved, emerge as key considerations to ethical design research (ibid).

Whilst these tools, guides and activities do provide frameworks and guidance for conducting ethical design research, they arguably fall short in identifying when it is/is not appropriate to involve users and experts and do not fully address the unpredictability of the design research and implementation processes. Therefore, there is a need for practical ways to respond to changes as they occur across the full project process as well as understanding when to limit user involvement whilst still yielding valuable insights and feedback.

2. Method

Seven product design research projects were chosen as exploratory case studies (Table 2). The projects are the work from within a school of design at (The University of Limerick)]. Cases were deliberately chosen across the various educational levels, from UG to PG to understand and identify challenges and opportunities at each level that could be both generalised or become specific to a particular project type and designer level. However, a certain level of design experience and standard of work was assumed from the design students, across all cases commensurate with the educational level of the student (e.g. UG or PG).

Table 2. Case studies.

Case Study	Project theme	Academic level	Participants	Participatory engagement and methods used	Alternative methods
C1	A multi-sensory interface for people with visual impairment	UG Degree project (6 months)	People with vision impairment, experts, simulated users, service providers	Mapping, empathy role plays, user interviews with discussion guide	Simulated testing with proxy users, gatekeeper interviews & testing
C2	Redesigning home birthing: a device to facilitate active labour during homebirths	UG Degree project (6 months)	People who have given birth, people who may give birth in the future, experts & clinicians in birthing process	Snowball recruitment, interviews & focus groups, expert interviews, user testing.	Forums & online blogs, simulated user testing with proxy users, retrospective interviews (past experiences)
C3	A non-invasive home testing kit for HPV using menstrual fluid	PG Master's project (9 months)	General population (fitting criteria), experts	Close-tie connections, survey through social media, scenario building, expert interviews, user interviews, user testing	Forums & blogs, willing participant for testing, proxy user testing, empathy testing, retrospective interviews
C4	An assistive device for older adults living at home	PG Master's project (9 months)	Older adults, experts	Purposive sampling,close-tie connections,. interviews & card sorting, user testing, experts	‘Tea & cake’ intro meeting, simulated testing with proxy users, gatekeeper interviews & testing, remote photo stories
C5	A vocal health device for patients of speech and language therapy (SLT)	PG Doctoral research (24 months)	Current users experiencing vocal issues, people who experienced vocal issues in past, general population (fitting criteria). experts in vocal performance & SLT	Purposive sampling, focus groups, iterative prototype testing, user testing, recruitment through gatekeeper	Empathy testing, simulated testing with proxy users, expert evaluation of prototypes
C6	An investigation and redesign of the speculum instrument	PG Master's project (9 months)	People who had experienced the issues previously (healthy users), experts	Anonymous survey,early-stage user testing	Artificial anatomy built for testing, ‘Show & tell’ interviews, expert evaluation of prototypes, simulated testing, retrospective interviews
C7	Mitigating injury in the design of the light shoe for Irish dancing	UG/PD Course project (2 months) Commercialisation	Users with prior injuries (students amateur & professional), experts (dance instructors, biomechanics experts), producers & suppliers	Interviews, performance reviews, immersive designer testing, recruitment through gatekeeper	Simulated testing with proxy users, expert evaluation of prototypes

The case studies, while all in the health and wellbeing space vary from incremental redesigns of existing solutions (e.g. Case 6: Speculum and Case 5: Vocal Health) to more radical explorations of challenges (e.g. Case 1: Interface for visual impairment and Case 2: Redesigning homebirth) and applications of new technologies to user challenges (Case 3: HPV home testing). This range served to demonstrate the breadth and depth of research and practice projects to which design-led approaches are applied. Within each case study, the intended users were classified as vulnerable, and in some or all cases, people (as users, experts or gatekeepers) were involved in the design process at different stages. The level of involvement varied across the projects due to access and availability of participants, perceived risk to participants, designer engagement and project direction and duration.

It is important to note that all activities undertaken in this research are underpinned by adherence to the HSE National Policy for Consent in Health and Social Care Research (2023) with oversight from the University of Limerick, Faculty of Science and Engineering Research Ethics Committee.

The case studies were retrospectively studied by two practitioner–researchers with over 30 combined years in design education and research. The main sources of data were the project process books, which document the entire project in both visual and written format. The process books were analysed, and this data was triangulated with notes from design tutors, who, in all cases, were also the researchers, taken during the projects. Through examining these cases, the researchers could explore the process undertaken by the designers for research and design development, the methods they used when working with participants across all project stages and the measures undertaken or overlooked, when working with these vulnerable participants.

3. Case examples

4. Findings and discussion

In the following section, we reflect on the findings from the case study analyses, which are briefly discussed and followed with proposed living guidelines and a preparation guide to support the process. Eight key themes emerged. Whilst some themes might be more appropriate to specific stages of the design process (e.g. Ethical Approval and Preparation align with primary research), some bridge across several if not all stages of the design process. Therefore, the order of discussion does not indicate an order by which the guidelines should be followed.

4.1 Ethical approval

It was clear from the variety of cases reviewed that there is no ‘one size fits all’ process for ethical design engagement with vulnerable users and that plans can change as design research evolves (Munteanu et al. 2015). Variables within the cases included: the duration of the project (shorter for UG projects- C1 and C2, longer duration for PG and PD C3 to C7); the direction the project took (conceptual systems/service approach- C1 and C4 or a concrete physical object as solution C6); the level of refinement the project reached (idea presentation – C1 and C3 through to ‘market release’ – C7) and the complexity and/nature of proposed solutions (risk associated with prototype testing, changing user needs).

It became clear through the analysis that standard, overly rigid ethical approvals need to be reviewed and revised to enable a staged approach where protocols can be revised as engagement, data collection and feedback loops shift and change. This was evidenced in several of the case studies where the involvement of participants needed to be considered on an ongoing basis. The initial recruitment of participants was in line with the ethics approval board. However, over the course of the project, it was necessary to reflect on when and where to engage with participants. For example, in project C1, it was anticipated at the outset of the project that participants would be asked to test iterative prototypes. On reflection with the student and tutor on the project, it was decided that it may not be appropriate to ask the vulnerable participants involved to repeatedly evaluate and test low-fidelity prototypes for a project that would most likely not be implemented. In this case, proxy users were used to test adapted solutions to roleplay the interaction of a visually impaired person. Actual targeted users were then asked to comment on more developed solutions.

As advocated by Frauenberger, Rauhala, and Fitzpatrick (2017), we propose an ethics guide that is reflexive and responsive to issues as they arise during the design research process. This must be embedded over the course of a project and transparent about the issues and decisions. While it may not always be possible to revert to the formal ethics board each time an ethical decision must be made, it will become necessary for tutors and designers to be educated to make these judgement calls.

All design research should of course undergo some form of ethical review at the outset. At a minimum, plans and protocols for how and when participants are involved should be prepared, project purpose and participant role(s) clearly explained, and consent forms signed by participants or guardians. In student UG projects, where interaction with vulnerable participants might be very minimal, design educators and project principal investigators (PI) can manage this process. These individuals and teams, therefore, would require ethics training to ensure they can make informed decisions on the risks involved. Ethical approval can, therefore, be a less procedural process with the tutors acting as the key decision-makers, with UG students gaining experience in ethical practice (Godbold, Lees, and Reay 2019).

Guidelines:

• Ethical training is essential for design educators and managers.

• For UG projects formal but ‘localised’ ethics approval that is managed locally, must be sought that outlines plans for participant involvement, highlights any potential risk and describes measures to overcome these.

• Dynamic and staged ethics approval processes for various project types.

4.2 Recruitment

Recruitment of potential users as participants proved difficult for student designers and novice design researchers and was further complicated when the topic was sensitive, and the potential users identified as vulnerable. The most common method across all cases was convenience sampling (Neuman 2007). Here, the designers drew on existing networks (personal and professional) to recruit users, experts, and other stakeholders and then asked these network connections to share any other contact details for other potential participants (snowball sampling). This is often more difficult for students as their own networks were not extensive, and certain participant groups were involved in multiple projects or over longer durations leading to participant fatigue (C7). Additional complications were evidenced when close-tie connections may have influenced the decisions made by the designers in a more subjective way as participant –researcher relationships strengthened (C3) (Munteanu et al. 2015). This could be mitigated or minimised by the introduction of critical friends or independent participants if project duration extends (Craig et al. 2021).

Guidelines:

• Close-tie participants should be balanced with objective or critical participants (critical friends) at key points in the design process (e.g. user testing & evaluation).

• Snowball recruitment is very effective for expert participants.

• Access to participants can be made through liaison and support groups or other stakeholders.

4.3 Engagement

Situational ethics acknowledges that consent may change over the process and that involvement may peak and wane as a project evolves. The level of involvement may also vary from limited participation to active codesign and shared decision-making. This was particularly evident in longer projects (C3 to C7) where designs are being brought to a higher level of development and opportunities for additional interactions arose, outside what was approved by the ethics board. Cases where there was a continuous dialogue between the stakeholders, ensured participants were aware of any changes and they could confirm that they were still willing to be involved in the project, were evidenced to be more productive in terms of process and outputs (C3, C5 and C7). Ongoing informed consent and clear information loops throughout are essential, so a participant is empowered to continue or withdraw at any stage of the process (HSE 2023). These are non-negotiables and where possible plain language and visuals should be used to avoid misunderstanding and information breaks.

Guidelines:

• Build in for additional activities in the ethics protocol from the outset (e.g. focus groups as well as individual interviews).

• Ongoing and dynamic consent throughout the process that is continuously monitored and reviewed throughout (Dewing 2008).

• Reconsent sought if the parameters of the project or the type and level of participatory activities change (HSE 2023).

• Language used is clear and understandable and ideally employs visual elements.

4.4 Gatekeepers

Direct access to vulnerable users was not always possible for the full project duration across all the cases examined (C2, C4 and C6). Where access was possible, it was typically through a gatekeeper who had an implicit understanding of the users, their needs, the culture and context and could advise on the most appropriate and sensitive way to engage with people. This person was often a professional service provider, an official advocate or a family member who could also offer insights as a key stakeholder in the area.

Other types of gatekeepers were charities, NGOs and patient support bodies where these services could provide guided access to participants (C1 and C2). Indeed, many of these gatekeepers were involved in later-stage testing and evaluation of concepts if it was not appropriate to involve users at these stages (C2, C4, C5 and C6). These gatekeepers were also cognizant of whether the participant cohort or individual has been involved in research previously so the research team could avoid overuse with a particular group.

Guidelines:

• Gatekeepers are essential for engagement with vulnerable participants. This protects both the participant and the designer.

• Gatekeepers can identify when it is/is not appropriate to involve specific participant cohorts or individuals as they have more complete understandings of participants’ needs and capabilities.

• Gatekeepers can have a deeper involvement in the design process providing expert feedback throughout.

• Caution needs to be exerted to avoid ‘over-using’ vulnerable participants across long-duration projects. Gatekeepers can monitor and minimise this occurrence.

4.5 Alternative participants

Mapping the key stakeholders and relationships at the start of most cases, helped designers to identify whether or where proxy or alternative participants could be used in place of vulnerable participants. In C2 and C6, the designers (as examples) conducted retrospective interviews with people who had experienced the situation but were no longer considered vulnerable (e.g. former patients). This proved a very useful method of conducting research, reviewing ideas and testing concepts as the risk to the person was eliminated but valuable insights and feedback were garnered.

Guidelines:

Alternative participants might include the following:

• Proxy users to step in for early-stage testing and evaluation.

• Participants who have experienced an issue but who are no longer vulnerable, for example, healthy adults who were patients previously.

• Liaison with support groups, charities

• Experts such as those providing services or care for the vulnerable participants can provide user insights where users themselves cannot.

4.6. Alternative methods

Like the employment of proxy users, the designers also identified situations where alternative research and testing methods could be used that did not require participant involvement thus avoiding ‘overuse’ or participant fatigue. In some cases (C1, C4 and C6), the designers independently assessed the need for the participants to be involved and made changes to their design plans accordingly. They used workarounds for research and user evaluation and testing that provided sufficient results to make decisions and move the project forward. In C6 the designer created an artificial anatomy to carry out simulated testing of solutions and combined this with ‘Show & Tell’ interviews and expert evaluation of prototypes. Whilst simulation will never fully replicate a real scenario, initial design issues can be resolved before returning to full user testing, where this is possible.

Participation of users is dynamic and must be reviewed at various stages of the process to avoid over-burdening participants whilst also recognising their interest in sustained involvement. A dynamic approach to participant involvement can overcome issues where access to the participants is restricted or in response to changes during the design process. This will also translate into a dynamic approach to ongoing consent where participants can opt out or change their roles. These steps are particularly important at UG where novice designers are at greater risk of mishandling sensitive or challenging situations. This to some degree goes against the principles of participatory (Cavignaux-Bros and Cristol 2020) and co-design (Dietrich et al. 2017) which advocate for the empowerment of participants by embedding them within the research and design process.

Guidelines:

• Workarounds should be the first choice where they yield the same results as participant involvement.

• Research workarounds: Expert & stakeholder primary research; online forums & blogs; empathy research; scenario Building; role play.

• Ideation & concept testing workarounds: proxy user testing; simulated testing; journey mapping; user stories. role play.

4.7. Feedback and giving back

The participant’s expectations were not always managed appropriately. Feedback loops were not always two-way, participation was not acknowledged and little in the way of payback to the participants was considered. For example, in C3, experts had participated in sharing their experiences in initial research and feedback on early ideas but were not updated on the final developed concept which led to them querying the progress of the project a year later. In this instance, sending all participants a visually engaging document outlining the developed concept would close out the project and acknowledge the participant’s input. Where the designer could ‘reward’ participants, buy-in was stronger, e.g. designer in C4 brought tea and cakes to the initial introductory meetings with users to build trust and ‘break the ice’.

Typically, the projects under study did not develop into fully realised designs, particularly those undertaken at UG (C1 and C2). At times the participants were not aware of this when they became initially involved in the project. This led to disappointment and at times disengagement from the process once the limits of the project were explained (C1).

While issues of Intellectual Property did not arise specifically in the cases studied, we became aware that there may be potential issues with intellectual property (IP). Particularly in longer projects with more sustained participant involvement (C7), some questions arose with the attribution of intellectual property as the ideas evolved and concretised. Identifying roles and then attributing IP accordingly in an equitable and transparent manner. This falls outside the scope of this article but poses some interesting questions for planning participatory projects in future research.

Guidelines:

• Designers must be explicit about the type of project being undertaken (UG college project) and manage the expectations of participants accordingly.

• Explore mechanisms for reward/ reimbursement in accordance with their time and effort input.

• Acknowledging the participant’s and gatekeeper’s input through continuous feedback loops demonstrates respect and can strengthen involvement and allow the process to evolve and develop in tandem with user’s needs and evolving situations.

4.8. Before the project begins

Along with the guidelines above, we have collated a list of questions, detailed in Table 3 to pose at the outset of a project. This will help decide where and if it is best to include vulnerable participants, how these engagements should then be handled with rigour and sensitivity and to what level this participation should reach. This set of questions and guide are living documents that should be added to and amended by design teams. The guidelines also allow for sustained, sensitive and equitable participation over a longer duration project. The guidelines also point to credible alternatives where access to participants isn’t available or advisable, thus, allowing design projects to proceed without dependencies on external stakeholders.

Table 3. Design research ethics guide.

Design has the capacity to uncover issues and problems and respond in an empathetic way to the needs of the people for whom the research is being conducted. Design research, unfortunately as with other disciplines, risks over-involving or over-using participants with little benefit to them in the longer or shorter term. It is hoped that if design researchers follow this set of guidelines as well as other similar methods and tools, that design can concrete its place as a credible and robust research practice. Other researchers in both design and related disciplines can have confidence in the efficacy of the work and apply the methods in similar areas.

5. Conclusion

The case studies chosen in this research demonstrate the variety of directions design research can take and the diversity of problems and challenges it can address. The lessons learned from these cases illustrate the range of issues at play and that there are several ethical considerations that must be understood before deciding to recruit vulnerable participants to design research. Design often differs from other research projects as participants may be involved not just in the research phase but over the full life cycle of the project to trial and test solutions. Even at the outset of a design-led project, it is difficult, if not impossible to determine the direction of the project as design research advocates a divergent approach (Council 2023). This can be even more difficult to determine for novice and early-career designers. Ethical knowledge and understanding, as with other design skills, need to be acquired early in design education and build incrementally through both UG and PG and onto professional practice or advanced research. In early UG, before any engagement with people occurs, design tutors can introduce ethical issues through projects focused on life-centred design, sustainability, and research methods. As designers’ competency increases so too can the complexity of the project briefs and the level of engagement with participants whilst following approved guidelines and ethics procedures. Thus, designers will inherently examine issues from various perspectives and foresee problems that may occur.

While responsible human-centred design research must involve the key stakeholders, including vulnerable ones, it may not always be ethical to do so. Our analysis prioritises a set of principles that can support ethical decision-making regarding the involvement of vulnerable users in design research and practice. We presented a ‘living’ guide that will hopefully start a broader and longer discussion on how and when we bring users, experts, and other key stakeholders into design research and what this involvement might look like to ensure all party’s needs are considered and valued across the entire process. We also presented alternatives methods where it may be inappropriate or difficult to recruit participants. We hope that our guide can provide a base for more systematic and deliberate engagement with ethics in future design research and practice. There are opportunities for further research to explore more deeply the role, experiences and perspective of the participants as they engage in design research projects as well as the opportunity to further review the rigour and validity of alternative methods for design research and testing.

Acknowledgments

The authors would like to thank the student designers, researchers, participants, gatekeepers, experts and co-designers involved in the case study projects. Support for several of the projects was provided through the University of Limerick Health Research Institute.

Disclosure statement

No potential conflict of interest was reported by the author(s).